1. Technical Field
The present exemplary embodiment relates to an image processing apparatus, a magnetic resonance imaging apparatus, and an image processing method for processing an image obtained by imaging an inside of a subject. Particularly, the present exemplary embodiment relates to an image processing apparatus, a magnetic resonance imaging apparatus, and an image processing method according to which a distribution of velocity variations of a body fluid can be faithfully imaged even if the velocity variations of the fluid are not cyclical, such as velocity variations of a Cerebrospinal Fluid (CSF).
2. Related Art
As a conventional method of observing dynamics of a body fluid, such as CSF or blood, by using a magnetic resonance imaging apparatus, there is a method called “a phase contrast method” or “a phase shift method” that uses a flow-encode gradient magnetic field (for example, see Matt A. Bernstein, Kevin F. King, and Xiaohong Joe Zhou, “Handbook of MRI Pulse Sequences”, Elsevier Academic Press, 2004, pp. 659-677).
FIG. 14 is a schematic diagram of a pulse sequence according to a conventional phase shift method. As shown in FIG. 14, usually, according to a phase shift method, a gradient echo method is used; and a phase shift proportional to a velocity is given to an image by applying a flow-encode gradient magnetic field (Gfe) (P3 shown in FIG. 14) between an excitation pulse (P1 shown in FIG. 14) and an echo signal (P2 shown in FIG. 14).
According to the phase shift method, it is basically assumed that subject velocity variations strongly correlate with an electrocardiogram waveform (ECG shown in FIG. 14). Precisely, as shown in FIG. 14, after a certain waiting time (Tdelay shown in FIG. 14) has elapsed since an R wave appearing on an electrocardiogram-gated waveform, a radio-frequency excitation is performed and then an echo signal is collected. To reconstruct one image, approximately 128 to 256 echo signals are usually required. Therefore, generally, according to the phase shift method, a procedure of collecting one echo signal with respect to each R wave is repeated 128 to 256 times by stepwisely changing a phase-encoding gradient magnetic-field pulse.
An image S(ra), which is obtained by performing reconstruction processing, such as a discrete Fourier transform, on the echo signals collected in this way, has a phase shift proportional to a velocity. A velocity image V(ra) indicating a distribution of velocities is obtained by performing processing on the image S(ra), for example, the processing being expressed by V(ra)=k(venc)·arg{S(ra)} (where, k(venc) denotes a proportional coefficient that changes in accordance with the shape of a pulse of the flow-encode gradient magnetic field), after performing correction processing because of imperfection of the apparatus or ununiformity of the static magnetic field.
However, when observing dynamics of a CSF by using the phase shift method, there are problems as described below.
For example, according to the phase shift method, an echo signal is collected with respect to each of a plurality of R waves, for example, 128 to 256 R waves, which are equivalent to time for approximately two-four minutes, and it is known that there is a fairly low correlation between velocity variations of an actual CSF and electrocardiogram gating. For this reason, the velocity of the CSF during collection of each echo signal fluctuates to a large extent, so that the velocity observed on a reconstructed image as a phase shift is just an average value of velocities while collecting echo signals.
Purposes of observing dynamics of a CSF vary, for example, there are a case of examining presence or absence of a CSF circulation, and a case of precisely examining whether traffic of a CSF is available into a space that seems closed at glance. In such case, to what extent the maximum velocity is, and how far a CSF in a certain portion reaches within a certain time are influential and meaningful, so that only an image of an average velocity in a relatively long time range obtained by the phase shift method is substantially insufficient. Moreover, according to the phase shift method, fluctuations in the velocity at the time of collecting each echo signal are large, consequently, ghost artifact tends to appear in the phase encoding direction on a reconstructed image, and the image quality has a limitation.
In addition to the phase shift method, there is a method of collecting an image within a relatively short time, such as one second, by using an Echo Planar Imaging (EPI) method, or a fast Spin Echo (SE) method (for example, see “Proc. of ISMRM (International Society of Magnetic Resonance in Medicine) 1998, No. 2138”). However, the method consumes time and efforts for reading images and is not practical, because images are not provided in a sorted manner for observing dynamics of a CSF, and a number of images have to be referred.
As described above, according to conventional methods, for example, the phase shift method, velocity variations of a body fluid that are not cyclical, such as velocity variations of a CSF, cannot be faithfully imaged.